Why Is the Sky Blue Experiment: Prove It in a Glass

Ask a classroom why the sky is blue and a third of the hands will say it reflects the ocean. That answer has survived for generations because the correct one sounds like a dodge until you can see it happen. This why is the sky blue experiment makes it visible in a litre of water: look at the container from the side and it has gone faintly blue, look along the beam and the far end has gone orange. One glass, one torch, a sky and a sunset at the same time. The payoff is a kid who can explain both, and who knows why the sky isn’t violet.

Key Takeaways

• You’ll scatter a white torch beam through water clouded with two drops of skim milk, then view it from two angles.
• It teaches Rayleigh scattering, why short wavelengths scatter more, and why the same physics makes sunsets red.
• Best from about age 7, and genuinely interesting to teenagers because of the violet problem below.
• What usually goes wrong: too much milk. Four drops in a litre is already too many, and whole milk ruins it.
• Honest limit: milk is an imperfect stand-in for air, and the article says exactly where the model breaks.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
Tall clear container	One, 1–2 litres, straight sides	Kitchen: a glass vase, a rectangular storage box, a tall jug	A 2-litre clear soft-drink bottle with the label peeled off
Water	Fill the container	Tap	—
Skim milk	2–6 drops total	Fridge	Powdered skim milk mixed very thin; whole milk works poorly (see below)
Bright white LED torch	One, 100 lumens or more	Drawer, or a phone torch	A phone torch works; a focused beam works better
Eye dropper or a drinking straw	One	Medicine cabinet	Dip a straw, cap the top with a finger, release one drop
Black card	One A4 sheet	Stationery drawer	A dark towel behind the container
Phone camera	One	Pocket	Optional, but it makes the result measurable

Allergen note: milk. If there is a dairy allergy in the house, a pinch of cornflour whisked into a cup of water first and then added a few drops at a time behaves similarly, though less cleanly.

Time: prep 5 min, experiment 20 min, cleanup 5 min Cost: everything is probably in your kitchen Mess level: low

Do the Experiment: The Why Is the Sky Blue Experiment, Step by Step

Step 1 — Fill the container and look through it

Fill with plain tap water and set it on black card in a dim room. Shine the torch in one end. Look from the side: nothing. Look along the beam: nothing. Clean water scatters almost no light, which is the baseline everyone skips and then wonders what changed.

Step 2 — Ask for the prediction before the milk goes in

What will happen when I add two drops of milk? Most children say “it will go white” or “cloudy”. Write the answer down. Then ask the harder one: will it look the same from the side as it does end-on?

Step 3 — Add exactly two drops of skim milk and stir once

Two drops. Not a splash. Stir gently with the straw so it disperses evenly, and give it ten seconds.

What you should see: the beam becomes visible as a shaft through the water, and viewed from the side the shaft has a faint blue-grey cast. It is subtle. That subtlety is the honest result, and a dark background plus a dark room is what makes it readable.

Step 4 — Walk round to the far end and look back along the beam

Stand so the beam is coming toward you through the water, with the torch pointed slightly off your eyeline rather than straight at your face.

What you should see: the light arriving at the far end has gone warm, yellow or faintly orange. The same water looks bluish from the side and orange from the end, at the same instant, from one white light source.

Step 5 — Add two more drops and watch the sunset deepen

Add two drops, stir, look end-on again. Then two more.

What you should see: the transmitted beam shifts from yellow toward orange and then a muddy red, while the side view loses its blue and goes milky white. You have just modelled the difference between noon and sunset, and then between a clear sky and a hazy one.

Step 6 — Photograph both views and read the numbers

Lock your camera’s exposure and white balance manually, then photograph the side view and the end view. Open both in a free image editor and read the average red, green and blue values in a fixed rectangle.

What you should see: in the side view, blue exceeds red. In the end view, red exceeds blue. The numbers are usually only 10 to 30 per cent apart, which is exactly the point. The sky’s blue is not a saturated colour either.

What’s Actually Going On

Air molecules are tiny compared with the wavelength of visible light. When light hits something that small, the scattered intensity depends very steeply on wavelength: it goes as one over wavelength to the fourth power. This is Rayleigh scattering, and that fourth power is the whole story (HyperPhysics, Georgia State University). Run the arithmetic with blue light at 450 nanometres and red at 650: the ratio of wavelengths is about 1.44, and raised to the fourth power that is roughly 4.4. Blue light is scattered out of a sunbeam something like four times more readily than red. Sunlight entering the atmosphere therefore sheds its short wavelengths sideways, which is what reaches your eye when you look away from the sun. The sky is lit by the light that got deflected.

The sunset is the same mechanism from the other end. At sunset you are looking along a much longer slant path through the atmosphere, so far more of the blue has been scattered out before the beam reaches you, and what survives is the red and orange remainder. Your glass of water does both at once because the side view shows you scattered light and the end view shows you transmitted light.

Now the question almost no article answers. If shorter wavelengths scatter more, violet scatters more than blue, so why isn’t the sky violet? Three reasons stack up. The sun emits less violet than blue to begin with. Ozone in the upper atmosphere absorbs part of what remains. And human colour vision is far less sensitive at 400 nanometres than at 450, so the violet that does arrive contributes little to what we perceive. The sky’s colour is the combined result of the scattering physics and the response of the eye looking at it, an argument worked through carefully by Glenn Smith in the American Journal of Physics (Smith, 2005). The sky is a desaturated pale blue, not a pure spectral blue, and the washed-out result in your container is closer to the truth than a cartoon sky.

Where the model breaks, honestly. Milk is not air. Its scattering comes mostly from casein micelles, protein clusters roughly 50 to 300 nanometres across, which are small enough to give wavelength-dependent scattering, and from fat globules between 0.1 and 10 micrometres, which are not. Large particles scatter all visible wavelengths about equally, which is the regime called Mie scattering and the reason clouds are white rather than blue. That is why skim milk works better than whole milk, and why two drops work better than ten. Add enough milk and you stop modelling a sky and start modelling a cloud.

How to Explain It to Your Kid

Ages 5–8: Blue light bounces more

Sunlight is all the colours mixed together. When it crashes into the air, the blue part bounces off in every direction, like a ball hitting a wall, and the red part mostly keeps going straight. So when you look up at a part of the sky away from the sun, you are seeing all that bounced-off blue. At sunset the light has to travel through much more air to reach you, so almost all the blue has bounced away somewhere else and only the red is left. Have them look at the glass from the side, then walk round and look at the end, and say out loud which colour each view is.

Ages 9–12: The fourth power, and why clouds are white

Introduce the rule: scattering depends on one over the wavelength to the fourth power. Do the sum together. Red light has a wavelength about 1.4 times longer than blue, and 1.4 multiplied by itself four times is about four, so blue scatters roughly four times as much. Then the twist that makes the rule real: this only holds when the scattering thing is much smaller than the wavelength. Cloud droplets are hundreds of times bigger, so they scatter every colour equally and come out white. That is why milk goes white when you add too much.

Ages 13+: Why the sky is not violet

Give them the full chain and let them argue with it. Scattering efficiency rises toward short wavelengths, but perceived colour depends on the solar spectrum, atmospheric absorption and the spectral sensitivity of the three cone types in the retina. Then set the real investigation: photograph the sky straight up, at 45 degrees, and just above the horizon, all with locked exposure. Plot the blue-to-red ratio against angle. The ratio falls as you approach the horizon, because the path length through the atmosphere grows, and that is the same curve their sunset is on.

The question to ask: “On the Moon there is no atmosphere. What colour is the sky at lunar noon, and why?”

Why Teaching This Early Pays Off

Children do not arrive with no theory of the sky. They arrive with one, usually that the sky is blue because it reflects the sea, and research on children’s mental models of the earth found that they hold such models coherently and patch them rather than abandon them when contradicted (Vosniadou & Brewer, 1992). That is the reason a demonstration beats a correction. Showing a child that one white light can produce blue in one direction and orange in another does not just add a fact. It makes the old model unable to account for what they are looking at.

The concept itself keeps paying. Wavelength-dependent scattering turns up again in why distant mountains look blue, why fog headlights are yellow, why astronomers worry about reddening, and in every physics course that covers waves. The Framework for K-12 Science Education calls the practice being trained here developing and using models, and it names explicitly the habit of knowing a model’s limits (National Research Council, 2012). The glass of milky water is a model with obvious limits, which makes it a better teaching object than a perfect one.

Before You Start: What Parents Should Know

Best age	7 and up for the two-view version. The violet argument lands from about 12
Supervision	Light. Water and a torch. The only adult job is rationing the milk
Real hazards	None beyond normal kitchen care. Don’t stare directly into a bright LED torch, and keep the container away from the table edge
Don’t	Don’t substitute whole milk, cream or coffee creamer when you want the blue. The fat globules are too large and you will get white every time. Powdered skim milk mixed very thin is the better substitute

When It Doesn’t Work

• Everything is white and nothing is blue. Too much milk, almost always. Start over with clean water and one drop. The useful range is narrow and sits lower than people expect.
• The blue is so faint you aren’t sure it’s there. Compare rather than judge. Set an identical container of plain water beside it and look at both through the same beam. Side-by-side makes a 15 per cent colour shift obvious where a single sample does not.
• The end view looks white, not orange. Your path length is too short. Use the longest container you have, pointing the beam down its length rather than across its width. A 2-litre bottle lying on its side gives about 30 cm of water, which is usually enough.
• A warm-tinted torch fakes the result. Many cheap LED torches are already yellowish, so the end view would look warm with no milk at all. Check the no-milk baseline first, and if the torch is warm, use your phone torch instead, which is usually closer to white.
• A teenager objects that milk is not air. They are right, and that objection is the best thing that will happen all afternoon. The casein and fat paragraph above is the answer, and conceding the point is more useful than defending the demo.

Take It Further

Angle: measure how the colour changes with viewing angle. Mark positions at 30, 60, 90 and 150 degrees around the container, photograph from each with locked exposure, and plot the blue-to-red ratio. Rayleigh scattering is not uniform in all directions, and the curve shows it.

Polarisation: scattered light is partly polarised, which is why a polarising filter darkens a blue sky. Hold one lens from a pair of polarised sunglasses in front of the camera and rotate it while photographing the side view. The brightness changes. The transmitted end view barely changes at all.

Where the Why Is the Sky Blue Experiment Sits in a Longer Sequence

This one needs a prerequisite to land well: a child who has already seen white light taken apart. Do dispersion with a prism indoors or build a spectroscope from a CD and a cereal box first, because “sunlight contains all the colours” has to be a thing they have witnessed rather than a sentence they have been told. Afterwards, the natural pair is why paint mixing and light mixing disagree, which attacks the other big colour misconception, and our broader collection of optics experiments for kids has somewhere to go next.

What to Watch For Over the Next 3 Months

• Week 4: at a sunset, they say something about the light travelling through more air. Usually garbled. Still correct.
• Month 2: they notice distant hills looking bluer than near ones and connect it, unprompted, to the same mechanism.
• Month 3: they ask a question the demo cannot settle, such as why the Moon’s sky is black or why Mars sunsets are blue. Both have real answers, and reaching for them is the sign the model took hold.

Frequently Asked Questions

Is the sky blue because it reflects the ocean?

No. The ocean’s colour is partly a reflection of the sky, not the other way round, and the sky is just as blue over Kansas and the Sahara as over the Pacific. The blue comes from sunlight scattering off air molecules, and the fourth-power wavelength dependence of that scattering is what favours blue.

Why is the sky not violet, if violet scatters even more?

Because the sun emits less violet than blue, ozone absorbs some of what is left, and human cones respond weakly near 400 nanometres. The perceived colour depends on the light, the atmosphere and the eye together, which is the argument made in detail by Smith (2005).

Why are clouds white if the sky is blue?

Cloud droplets are hundreds of times larger than the wavelength of visible light, so they scatter all colours roughly equally. Different size regime, different rule. That is also what happens in your container once you add too much milk.

Can I use a laser pointer instead of a torch?

Don’t. A laser emits one wavelength, so it cannot show a colour difference, and pointing one through water adds eye risk for no gain. A white LED is the right tool here precisely because it contains all the colours.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Nave, R. “Blue Sky and Rayleigh Scattering.” HyperPhysics, Georgia State University. http://hyperphysics.phy-astr.gsu.edu/hbase/atmos/blusky.html
2. Smith, G. S. (2005). “Human color vision and the unsaturated blue color of the daytime sky.” American Journal of Physics, 73(7), 590–597. https://doi.org/10.1119/1.1858479
3. NASA Science. “Visible Light.” Introduction to the Electromagnetic Spectrum. https://science.nasa.gov/ems/09_visiblelight/
4. Vosniadou, S., & Brewer, W. F. (1992). “Mental models of the earth: A study of conceptual change in childhood.” Cognitive Psychology, 24(4), 535–585. https://doi.org/10.1016/0010-0285(92)90018-W
5. National Research Council. (2012). A Framework for K-12 Science Education. National Academies Press. https://doi.org/10.17226/13165
6. White, R. T., & Gunstone, R. F. (1992). Probing Understanding. Falmer Press.